Avisos de código aberto
You are viewing the English version of this document. English is the authoritative version.
PodcastAI uses open-source software and models made by others. This page lists the ones we use in our product, who made them, the license each one is under, and where it comes from.
Last checked: October 16, 2026
| Component | What We Use It For | Made By | License | Source |
|---|---|---|---|---|
| YuNet (face_detection_yunet_2023mar) | Finding faces in video, so Studio can frame the person speaking | Shiqi Yu | MIT | OpenCV Zoo |
| SFace (face_recognition_sface_2021dec) | Telling whether two faces belong to the same person | Apache License 2.0 | OpenCV Zoo | |
| Light-ASD | Telling who is speaking, from their lips and the sound | Liao and others (CVPR 2023) | MIT | |
| AVA-ActiveSpeaker | The data Light-ASD's model was trained on | CC BY 4.0 | ||
| Caption and title fonts, including Inter | The fonts you choose for captions and titles in Studio | Google Fonts |
Blank cells mean a detail is not yet listed here.
We update this page and its date when the software and models we use change.